Cultural Networks and Bridges: How Advocacy Organizations Stimulate Viral Social Media Conversations

Chris Bail
Duke Sociology

Economic Consequences of ASDs

Explanations for the rise of Autism Spectrum Disorders (ASDs)

  1. Vaccines
  2. “Cold” Mothers
  3. Lyme Disease
  4. Relative Finger Length
  5. Wealth
  6. Democratic Party Membership
  7. Pesticides
  8. Air Pollution
  9. Androgen
  10. The Internet

Social Media and Autism

How do Advocacy Organizations Stimulate Conversation about Autism Spectrum Disorders on Social Media?

Outline of Talk

 

Outline of Talk

 

  1. A new theory of cultural resonance

Outline of Talk

 

  1. A new theory of cultural resonance

  2. Mapping culture using natural language processing and network analysis
     

Outline of Talk

 

  1. A new theory of cultural resonance

  2. Mapping culture using natural language processing and network analysis

  3. Using social media applications for social science research

Outline of Talk

 

  1. A new theory of cultural resonance

  2. Mapping culture using natural language processing and network analysis

  3. Using social media applications for social science research

  4. Results/Conclusions

What is cultural resonance?

Problems with the concept of cultural resonance

Problems with the concept of cultural resonance

 
1) A concept instead of a theory?

Problems with the concept of cultural resonance

 
1) A concept instead of a theory?

2) Idiosyncratic

Problems with the concept of cultural resonance

 
1) A concept instead of a theory?

2) Idiosyncratic

3) Circular reasoning

Resonance as Cultural Bridging

A One Day Slice of the Cultural Network

Hypothesis

 

Advocacy organizations whose social media messages create cultural bridges– or link topics of discussion that are seldom discussed together– are more likely to stimulate comments from new audiences. This is not only because they appear both new and familiar to multiple audiences, but also because they put these audiences into conversation with each other.

Resonance as Cultural Bridging

Research Design

Research Design

  1. Identified sample of Autism advocacy organizations via IRS database/Facebook searches
     

Research Design

  1. Identified sample of Autism advocacy organizations via IRS database/Facebook searches
  2. Extracted all posts produced by organizations as well as comments about them from Facebook's Application Programming Interface (API) for the period between Dec 31, 2011 to January 1 2013

Research Design

  1. Identified sample of Autism advocacy organizations via IRS database/Facebook searches
  2. Extracted all posts produced by organizations as well as comments about them from Facebook's Application Programming Interface (API) for the period between Dec 31, 2012 to January 1 2013
  3. Counted number of unique social media users who made comments of more than three words about each organization's posts by day who had not previously commented upon any of the organization's posts.  

Research Design

  1. Identified sample of Autism advocacy organizations via IRS database/Facebook searches
  2. Extracted all posts produced by organizations as well as comments about them from Facebook's Application Programming Interface (API) for the period between Dec 31, 2012 to January 1 2013
  3. Counted number of unique social media users who made comments of more than three words about each organization's posts by day who had not previously commented upon any of the organization's posts.
  4. Mapped cultural networks using technique using NLP and network analysis techniques; calculated “cultural betweenness” of each organization by day  

Constructing Cultural Networks

Calculating Cultural Betweenness

 
Modified Djikstra Algorithm for weighted networks:

Cultural Betweenness of each organization (i) by day (t):

The Cultural Network of ASD Advocacy Orgs on Facebook

The Cultural Network of ASD Advocacy Orgs on Facebook

Outcome: Number of Substantial Comments from New Audiences

...But what about everything else?

 

...But what about everything else?

 

  1. The wealth and size of an organization;
     

...But what about everything else?

 

  1. The wealth and size of an organization
     
  2. Social networks (instead of cultural networks)

...But what about everything else?

 

  1. The wealth and size of an organization
     
  2. Social networks (instead of cultural networks)
     
  3. The size and characteristics of its audience
     

...But what about everything else?

 

  1. The wealth and size of an organization
     
  2. Social networks (instead of cultural networks)
     
  3. The size and characteristics of its audience
     
  4. Broader factors (e.g. news or blog coverage, high profile events)
     

There's an App for that!

Data Collection Process

 

Response Rate

 

  1. 33.81% of all organizations contacted installed the app.

Response Rate

 

  1. 33.81% of all organizations contacted installed the app.
  2. Minimal evidence of selection bias.

Results

Interpretation

 

Organizations whose messages create substantial cultural bridges receive approximately 2.52 times more substantial comments from new audiences than those that do not

Limitations

 

Limitations

 
1. Causal inference is not possible

Limitations

 
1. Causal inference is not possible
 
2. Analysis limited to public Facebook pages

Limitations

 
1. Causal inference is not possible
 
2. Analysis limited to public Facebook pages
 
3. No analysis of sentiment/valence of messages

Limitations

 
1. Causal inference is not possible
 
2. Analysis limited to public Facebook pages
 
3. No analysis of sentiment/valence of messages
 
4. Unable to determine whether online processes have offline consequences

Conclusion

 

Conclusion

 

  1. New theory of cultural resonance
     

Conclusion

 

  1. New theory of cultural resonance
     
  2. New tools for measuring culture
     

Conclusion

 

  1. New theory of cultural resonance
     
  2. New tools for measuring culture
     
  3. New research design that uses social media applications in order to recover the “social science” within “computational social science”  

TextNets: A New R Package

http://github.com/cbail/textnets

More TextNets (Taylor Brown, Duke)

Summer Institute in Computational Social Science

 

Comments Welcome.

 

www.chrisbail.net
christopher.bail@duke.edu
@chris_bail  

Paper available here:
http://bit.ly/2dQpp0g

Generalized Additive Model

Emotional Feedback

Emotional/Cognitive Rhythms

Examining different Lag Lengths

 

Cultural Bridges in LGBT Advocacy

Cultural Bridges in Climate Change Advocacy

Cultural Bridges in Organ Donation Advocacy